No abstract

No abstract

Printed on paper containing at least 50% wastepaper, including 10% post consumer waste. iv AcknowledgmentsThe acoustic array design and testing team would like to acknowledge Scott Wilde and his operations team for their support in setting up, testing, and maintaining our equipment. Arlinda Huskey's help in supporting the program's logistics is much appreciated. Andy Scholbrock has been invaluable in coordinating availability of the Controls Advanced Research Turbine (CART)-2. Mike Asheim's input in array design, construction, and associated acoustic testing was also very useful. Finally, the project would not have been possible without the cooperation of the University of Colorado Boulder, their facilities, and the education they provide. Executive SummaryA prototype acoustic array for measurement of wind turbine noise has been designed in a collaborative effort by the National Renewable Energy Laboratory (NREL) and University of Colorado Boulder. The array is built upon 63 electret condenser microphones in a multi-arm logarithmic spiral pattern, a single calibrated B&K precision condenser microphone, in-house designed analog electronics, and an integrated National Instruments (NI) data acquisition system (DAS). Noise source maps are generated using a delay-and-sum beamforming algorithm with diagonal elimination in the cross-spectral matrix. The uniqueness of this array design over currently available commercial products is a lower frequency range, a weather-proof design allowing for long-term observational campaigns, and the ability to collapse and package the array for ease of transport.After an initial deployment at the Northwind 100A research turbine site for system testing, the array was deployed in late 2012 at the upwind side of the Controls Advanced Research Turbine (CART)-2 at the National Wind Technology Center (NWTC). The turbine has a highly instrumented rotor and accompanying meteorological tower enabling noise correlations with both detailed turbine operational characteristics and atmospheric conditions. An active acoustic source has been used to calibrate the array and characterize its performance. Simulated pointsource data have been used to compare these results to expected results. The array is shown to have a 3-dB beamwidth of 3.53 meters for a 750-Hz source at the center of the rotor, 56 meters from the center of the array. Turbine noise data have been collected over a range of operational conditions, including wind speeds from 7 to 22 m/s and rotor speeds of up to 42 rpm (6-second averages). Data from these measurement campaigns have been analyzed at frequencies from 750 Hz to 8 kHz.The array in its current configuration has shown point-source localization accuracy on the order of a few tens of centimeters. However, a non-negligible variation in source location under various atmospheric conditions has also been observed. Compensation for sound convection using a simplified atmospheric model in the beam-forming algorithm improved the initial results but only approximately half of the positi...

No abstract